Motor Fault Diagnosis Based on D-S Evidence Theory Information Fusion Algorithm

Abstract

D-S evidence theory is an uncertain reasoning method. This paper firstly studies the information fusion algorithm based on D-S evidence theory, including the combination rules of D-S evidence theory and the fusion process of fault diagnosis information. Then the common motor faults are summarized into bearing faults, stator faults, rotor faults and eccentric faults. In the experimental analysis part, four independent sensors are used to detect motor faults. After D-S evidence theory information fusion, the confidence distribution value of faults increases correspondingly with each sensor information fusion. This shows that more sensors can improve the accuracy of motor fault diagnosis results. The innovation lies in assigning basic credibility to the evidence, using combination rules to obtain new basic credibility, and finally making decisions to obtain diagnostic results. The results show that D-S evidence theory has become an important method to deal with uncertain, incomplete and inaccurate information in the field of information fusion.